Code Warnings: Attribute undefined: .float32 of np: numpy module

edit

I use the following code snippet in my script:

   import numpy as np
   ...
   eig=np.zeros((nks,nbnd),dtype=np.float32)

In this case, wing will give the Code Warnings:

Attribute undefined: .float32 of np: numpy module

But in fact, this is the correct usage for numpy. Any hints for this problem?